Depression as an Evolutionary Strategy for Defense Against Infection

8/20/2019 Depression as an Evolutionary Strategy for Defense Against Infection

1/14

Review

Depression as an evolutionary strategy for defense against infection

Sherry Anders a, Midori Tanaka b, Dennis K. Kinney c,d,⇑

a Clinical Psychologist in Independent Practice, Boxborough, MA, USAb Obara Hospital, Sapporo, Hokkaido, Japanc McLean Hospital, Belmont, MA, USAd Department of Psychiatry, Harvard Medical School, Boston, MA, USA

a r t i c l e i n f o

Article history:

Available online 20 December 2012

Keywords:

Depression

Evolution

Mood

Inflammation

Infection

Infection-defense

Immune

Hypothesis

Genetic

Anti-depressants

a b s t r a c t

Recent discoveries relating depression to inflammation and immune function may help to solve an

important evolutionary puzzle: If depression carries with it so many negative consequences, including

notable costs to survival and reproduction, then why is it common and heritable? What countervailing

force or compensatory advantage has allowed susceptibility genes for depression to persist in the popu-

lation at such high rates? A priori, compensatory advantages in combating infection are a promising can-

didate, given that infection has been the major cause of mortality throughout human history. Emerging

evidence on deeply rooted bidirectional pathways of communication between the nervous and immune

systems further supports this notion. Here we present an updated review of the ‘‘infection-defense

hypothesis’’ of depression, which proposes that moods—with their ability to orchestrate a wide array

of physical and behavioral responses—have played an adaptive role throughout human history by helping

individuals fight existing infections, as well as helping both individuals and their kin avoid new ones. We

discuss new evidence that supports several key predictions derived from the hypothesis, and compare it

with other major evolutionary theories of depression. Specifically, we discuss how the infection-defense

hypothesis helps to explain emerging data on psychoimmunological features of depression, as well as

depression’s associations with a diverse array of conditions and illnesses—including nutritional deficien-

cies, seasonal changes, hormonal fluctuations, and chronic diseases—that previous evolutionary theoriesof depression have not accounted for. Finally, we note the potential implications of the hypothesis for the

treatment and prevention of depression.

2012 Elsevier Inc. All rights reserved.

1. Introduction: rationale for an ‘‘infection-defense’’ hypothesis

of depression

Infection has been the leading cause of mortality throughout

human history (Cairns, 1997; Finch, 2010). It has been estimated

that prior to the industrial period, the average life expectancy

was 25, and it was not uncommon for half of the siblings in a fam-

ily to die before reaching adulthood (Cairns, 1997; Casanova and

Abel, 2005). Particularly virulent pathogens could wipe out an en-tire family or village, such as the English ‘‘sweating sickness’’

known to have wiped out one-half to two-thirds of the population

in many English towns during the late 1400s and early 1500s

(Thwaites et al., 1997). With such stark capabilities, infection has

been a critical and potent driving force in natural selection

(Dobson and Carper, 1996). Specific alleles have evolved in re-

sponse to common pathogens in an environment; however, patho-

gens are ubiquitous and wide-ranging, with new forms continually

evolving, leaving individuals intrinsically vulnerable (Casanova

and Abel, 2005; Dobson and Carper, 1996). The ideal system of de-

fense against this inherent vulnerability to infection requires a

generalized response that is proactive in reducing infection risk

during times of increased vulnerability, as well as both flexible

and adaptive enough to provide resistance to a wide range of

pathogens.

Here, and in several recent papers (Kinney and Tanaka, 2009;

Tanaka and Kinney, 2011a,b; Tanaka et al., 2012), we propose thatmoods—with their ability to orchestrate a wide array of physical

and behavioral responses—have evolved as part of a complex sys-

tem of immune defense that helps counteract our inherent vulner-

ability to the diversity of environmental pathogens. The ‘‘infection-

defense hypothesis’’ offers a novel evolutionary framework for

understanding how many of the social and behavioral features of

depression may help individuals fight existing infections, as well

as help both individuals and their family members avoid new ones.

In contrast to many previous evolutionary theories of depression, it

takes into account and helps to integrate a large and growing body

of evidence linking depression to inflammation and immune func-

tion, and helps to explain depression’s association with a vast array

0889-1591/$ - see front matter 2012 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.bbi.2012.12.002

⇑ Corresponding author. Address: 18 Locust Avenue, Lexington, MA 02421, USA

Tel.: +1 781 862 1644, mobile: +1 617 271 5156; fax: +1 781 862 6559.

E-mail address: [email protected] (D.K. Kinney).

Brain, Behavior, and Immunity 31 (2013) 9–22

Contents lists available at SciVerse ScienceDirect

Brain, Behavior, and Immunity

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y b r b i

http://dx.doi.org/10.1016/j.bbi.2012.12.002mailto:[email protected]://dx.doi.org/10.1016/j.bbi.2012.12.002http://www.sciencedirect.com/science/journal/08891591http://www.elsevier.com/locate/ybrbihttp://www.elsevier.com/locate/ybrbihttp://www.sciencedirect.com/science/journal/08891591http://dx.doi.org/10.1016/j.bbi.2012.12.002mailto:[email protected]://dx.doi.org/10.1016/j.bbi.2012.12.002


2/14

of conditions and illnesses such as nutritional deficiencies, sea-

sonal changes, hormonal fluctuations, and chronic diseases. The

infection-defense hypothesis may also help to resolve a baffling

evolutionary puzzle—why major psychological depression is so

prevalent and heritable, despite its high costs for survival and

reproduction, including suicide (Tanaka and Kinney, 2011a).

We begin the paper with an overview of the epidemiology and

neurobiology of depression, including recent updates linking

depression to immune factors. This is followed by a discussion of

the infection-defense hypothesis (Kinney and Tanaka, 2009), in

which we propose that depressive features provide advantages in

combating infectious diseases—advantages that offset many

known disadvantages of depression. We also briefly review other

hypotheses and evidence that have related changes in immune

function to depression. We then describe several testable predic-

tions of the ‘‘infection-defense hypothesis’’ and discuss empirical

evidence that bears on each prediction. We conclude by comparing

the hypothesis with other evolutionary theories of depression and

by discussing some potential implications of the hypothesis for

better treatment and prevention of depression.

2. Depression: an evolutionary puzzle

While depression can occur at any point during the lifespan, its

most frequent onset occurs during the peak years of work and

reproduction—and both depressed individuals and their family

members often endure serious physical, social, and economic bur-

dens as a result (Broadhead et al., 1990; Eaton et al., 1997; Kler-

man, 1989). Depression negatively impacts productivity, as well

as psychosocial functioning, and is associated with increased rates

of unemployment and divorce (Weissman et al., 1996). Depression

is associated with lower rates of fertility (Tondo et al., 2011; Wil-

liams et al., 2007; Yates et al., 2010), and children of depressed

mothers show poorer average outcomes on a wide range of devel-

opmental indices (Cummings and Davies, 1994), with adverse con-

sequences noted even in cases where there has only been prenatalexposure to maternal depression (Davis et al., 2007). Increased

mortality rates are also associated with depression, as depression

is a risk-factor for many disease-related causes of death as well

as for suicide (Mykletun et al., 2007). From a global perspective,

the World Health Organization (WHO) has made the projection

that, by the year 2020, depression will be the 2nd leading cause

of disease burden worldwide (Murray and Lopez, 1996).

Depression thus poses a baffling evolutionary puzzle; despite

the serious consequences of depression for individuals and their

family members, including decreased fertility and increased mor-

tality rates, depression remains both common and heritable. The

estimated lifetime risk of a major depressive episode has risen to

23% in the United States (Kessler et al., 2005), and there is evidence

to suggest that the incidence of major depressive disorder mayactually be increasing (e.g., Compton et al., 2006). Moreover, the

heritability of depression is well-established, with estimates based

on twin, adoption, and genetic molecular studies consistently fall-

ing at about 40% (Kendler et al., 1995; McGuffin et al., 1996; Shyn

and Hamilton, 2010; Sullivan et al., 2000; Wender et al., 1986). Ge-

netic association and linkage studies have begun to discover spe-

cific alleles that increase risk for depression (see review by

Goldberg, 2006), such the CYP2C9⁄3 allelle (LLerená et al., 2003)

and the 5-HTTLPR short allele of the serotonin transporter gene

(Eley et al., 2004; Kendler et al., 2005).

For more than half a century, efforts to understand the neuro-

biological underpinnings of depression have been dominated by

the view that depression is caused by a deficiency in synaptic con-

centrations of monoaminergic neurotransmitters including seroto-nin and norepinephrine (Hirschfeld, 2000; Schildkraut and Kety,

1967). This idea, known as the monoamine hypothesis, has stimu-

lated a wealth of research and has been the major driving force be-

hind antidepressant drug development. Over time, however, the

initial promise of the monoamine hypothesis has been tempered

by the fact that attempts to find direct links between monoaminer-

gic transmission and mood have yielded equivocal results (Delgad-

o, 2000; Heninger et al., 1996). In addition, the efficacy of

antidepressant drugs based on the fundamental premise of the

monoamine hypothesis has been limited, with estimates that be-

tween 30% and 50% of individuals treated with antidepressant

medication do not show adequate response (Schatzberg, 2000).

The insufficiency of the monoamine hypothesis to explain critical

aspects of mood regulation and the desire for more favorable treat-

ment outcomes have resulted in an expanded search for the neuro-

biological underpinnings of depression.

Research on the neuroimmune system and its role in the etiol-

ogy of depression has emerged as an especially promising area for

study. In particular, the role of immune-activated inflammatory

cytokines has been identified as a key area of focus in understand-

ing the neurobiological pathways that trigger depressive states, by

way of direct and indirect effects on hypothalamic–pituitary–adre-

nal (HPA) axis, and by altering monoamine neurotransmitters in

multiple regions of the brain (Dantzer et al., 2008; Loftis et al.,

2010; Raison et al., 2006). In addition, a recent review of risk alleles

for depression has revealed that in a striking majority of cases the

depression alleles were associated with known effects on immune

function (Raison and Miller, 2012). These new lines of research—

which show wide ranging links between depressive symptomatol-

ogy and immune function—not only have the potential to lead to

novel treatment strategies for the prevention and treatment of

depression, but may also provide important clues as to the reasons

for depression’s prevalence and persistence throughout human

history.

3. Immune alterations, mood, and the macrophage and

cytokine theories of depression

Numerous associations between depression and immune func-

tion have been observed in recent years. Early studies investigating

immune alterations during depression focused almost exclusively

on markers of suppressed cellular immunity—such as decreased

lymphocyte proliferation and natural killer (NK) cell activity (Rei-

che et al., 2004; Schleifer et al., 1989; Zorilla et al., 2001). More re-

cently, however, there has been a shift toward understanding the

role of inflammation in depression, with a particular focus on the

role of proinflammatory cytokines (Zorilla et al., 2001; Irwin and

Miller, 2007). Increased levels of proinflammatory cytokines—such

as interleukin-1b (IL-1b), interleukin-6 (IL-6), tumor necrosis fac-

tor-a (TNF-a), and interferon-c (IFN-c)—have been repeatedly ob-served in depressed individuals, prompting formulation of the

‘macrophage theory of depression’ (Smith, 1991), and its succes-

sive formulation as the ‘cytokine hypothesis of depression’ (Maes

et al., 1995; Maes, 1999; Raison et al., 2006; Schiepers et al.,

2005). According to the cytokine hypothesis, proinflammatory

cytokines produced by macrophages during the acute phase of an

immune response act as neuromodulators that mediate the behav-

ioral and neurobiological features of depression.

3.1. Cytokine-induced changes in somatic experience, cognition, and

behavior

When infection or injury occurs, proinflammatory cytokines are

responsible for orchestrating the early immune response, including

sickness behavior. Sickness behavior—characterized by somatic,cognitive, and behavioral changes, such as fever, weakness, mal-

10 S. Anders et al. / Brain, Behavior, and Immunity 31 (2013) 9–22


3/14

aise, listlessness, hyperalgesia, and impaired concentration (Hart,

1988)—represents an organized strategy for fighting infection by

conserving metabolic resources and helping an individual avoid

further stressors (Hart, 1988; Kluger, 1991; Segerstrom, 2010; Yir-

miya et al., 2000). Dantzer (2001, 2009) has extended this notion to

suggest that sickness behavior is an expression of a biologically-

mediated motivational state triggered by the innate immune sys-

tem that resets an organism’s priorities to adaptively cope with

the threat of bodily insult. The costs of shifting resources and pri-

orities during this state are purportedly offset by the critical advan-

tages offered for fighting infection.

A portion of individuals who experience sickness behavior go on

to develop major depressive disorder. There is evidence that sec-

ondary development of depression following sickness behavior oc-

curs in individuals who have an exaggerated vulnerability to

infection (Dantzer, 2009; Wichers et al., 2006). For example, the

therapeutic administration of cytokines such as interferon-a(IFN-a) and interleukin-2 (IL-2)—used to treat patients with cancerand a number of infectious diseases—has been found to induce a

two-phase response. First, a few days following cytokine adminis-

tration, many patients experience sickness behavior, marked by

neurovegetative and somatic symptoms (fatigue, aches, loss of

appetite, sleep disturbance). Second, up to 50% of those treated

go on to develop symptoms of major depression, including de-

pressed mood, feelings of worthlessness, guilt, and even suicidal

ideation (Capuron et al., 2002; Capuron and Miller, 2004). Individ-

uals who are most likely to develop depression during the course of

cytokine therapy have been found to show increased baseline lev-

els of proinflammatory cytokines prior to treatment (Wichers et al.,

2006), and more severe levels of sickness-type behavior following

initial administration of treatment (Robaeys et al., 2007; Wichers

et al., 2005). Notably, cytokine-induced symptoms of depression

are attenuated by pretreatment with antidepressant therapy

(Capuron et al., 2002).

3.2. Sickness behavior and depression compared

There are many parallels between sickness behavior and

depression, although the two conditions are not equivalent (see Ta-

ble 1). For example, similar to sickness behavior, symptoms of

depression such as anhedonia, fatigue, hypersomnia, and psycho-

motor retardation (i.e., slowed speech, thinking, and body move-

ments) all tend to reduce activity and encourage rest, thereby

conserving energy. Some studies have found that depression is

associated with mild elevations in body temperature (e.g., Avery

et al., 1999; Rausch et al., 2003), and there is suggestive evidence

that milder temperature elevations, particularly when combined

with reduced bodily iron stores, may be inhibitive to pathogen

growth (Ismael and Bedell, 1986; Kluger and Rothenburg, 1979).

Nevertheless, a major difference between sickness behavior and

depression is that the costly antibiotic strategy of fever is notably

subverted during depression (Maier and Watkins, 1998). Depres-

sion also differs from sickness behavior in that its symptoms are

more variable and include behavioral and cognitive features that

discourage social contact and activity. Several authors have sug-

gested that depressive symptoms, like sickness behavior, may not

just be a spurious byproduct of cytokine-induced inflammation,

but may constitute an adaptive longer-term strategy for fighting

infection (Raison et al., 2006; Kinney and Tanaka, 2009).

3.3. Neuromodulatory mechanisms of cytokines

Multiple pathways have been identified by which cytokines ex-

ert neuromodulatory effects, and these have been covered exten-

sively in a number of recent reviews (e.g., Capuron and Miller,

2011; Dantzer et al., 2008; Maes et al., 2009; Miller et al., 2009;

Loftis et al., 2010). Proinflammatory cytokines have been noted,for example, to exert both direct and indirect effects on mono-

amine neurotransmitter availability and metabolism (Dunn and

Wang, 1995), as well as to increase activation of the hypotha-

lamic–pituitary–adrenal (HPA) axis (Holsboer, 2000; Pace et al.,

2007). A number of recent studies have also demonstrated links

between inflammatory activation and brain changes that underlie

depression (e.g., Brydon et al., 2008; Eisenberger et al., 2010; Har-

rison et al., 2009).

Cytokines administered to laboratory animals and humans have

been shown to alter the metabolism of the neurotransmitters sero-

tonin, norepinephrine, and dopamine in areas of the brain that are

associated with mood regulation (Rivier et al., 1989; Shuto et al.,

1997; Capuron et al., 2003). A number of studies suggest that cyto-

kines may decrease the amount of serotonin available for neuro-transmission by up-regulating the expression and activity of the

serotonin transporter (SERT) (Katafuchi et al., 2006; Zhu et al.,

2006). One of the most studied pathways by which cytokines

may influence neurotransmitter metabolism, however, involves

degradation of tryptophan by activation of the enzyme, indole-

amine 2,3-dioxygenase (IDO) (Capuron et al., 2003; Dantzer

et al., 2008). IDO is activated by proinflammatory cytokines

released during the acute-phase immune response, including

IFN-a, IL-6, and TNF. Serotonin levels are reduced due to the

Table 1

A comparison of sickness behavior and depression based on the ‘‘infection-defense hypothesis’’.

Sickness behavior Depression

Eliciting Conditions Infection, trauma, injury Immune vulnerability, immune compromise

Mode of Action Reactive Proactive and/or complementary; may act preemptively to avoid infection

during times of increased immune vulnerability, or help combat existing infections

Symptoms Circumscribed, including fever, malaise, muscle

and joint aches, impaired concentration, fatigue,

loss of appetite

Variable, includingsadness, loss of interest and pleasure, lowmotivation, withdrawal,

impaired concentration, fatigue, decreased sex drive, decreased speech, psychomotor

slowing, appetite disturbance, sleep disturbance, low self- esteem, hopelessness,

guilt, suicidal ideation, and somatic changes such as increased pain sensitivity, aches,

and mild hyperthermia.

Depressive symptoms are diverse and vary across patients and subtypes (e.g.,

melancholia may involve hypervigilance with agitation and insomnia that is less

common in atypical, postpartum, or seasonal subtypes); however, nearly all

symptoms appear to help fight or avoid infection in some way.

Duration Acute; sustained activation is biologically

costly and harmful to host

Sustainable for indefinite periods of time

Inflammatory Response Sickness behavior always involves inflammation Depression commonly, but not necessarily, involves inflammation

Adaptive Function Acute mobilization of resources to combat

infection and promote healing

Help combat existing infections, help avoid new immune-compromising stressors,

and help individuals and their kin avoid new infections

S. Anders et al. / Brain, Behavior, and Immunity 31 (2013) 9–22 11


4/14

breakdown of tryptophan by IDO, resulting in the production of

several neurotoxic compounds that may further mediate a depres-

sive response (Dantzer et al., 2008). IDO activity diverts tryptophan

metabolism from the production of serotonin to the synthesis of its

primary metabolite kynurenine, which is further metabolized into

several neuroactive compounds that include 3-hydroxykynurenine

(3-HK) and quinolinic acid (QUIN) or kynurenic acid (KA). These

compounds, in turn, generate free radicals that cause neuronal

damage due to oxidative stress (Wichers and Maes, 2004). The po-

tential importance of kynurenine metabolites in inflammatory-

mediated depression is underscored by the results of a recent

study in which Raison et al. (2010) found that individuals receiving

IFN-a for treatment of Hepatitis C were found to have higher cere-brospinal fluid (CSF) levels of kynurenine and its metabolites QUIN

and KA, and that these increases were correlated with increased

inflammatory biomarkers as well as depression.

Depression also has well-established links with hyperactivity of

the HPA-axis (Pariante and Lightman, 2008). Cytokines may con-

tribute to depression either directly, via activation of the HPA-axis,

or indirectly, through cytokine-induced glucocorticoid-receptor

resistance (Holsboer, 2000; Pace et al., 2007). Cytokines activate

the HPA-axis by inducing corticotropin-releasing hormone (CRH)

and vasopressin (AVP)—key regulators of the HPA-axis that are

found at increased levels in depressed individuals (see Owens

and Nemeroff, 1991; Holsboer, 2000; Scott and Dinan, 2002). Cyto-

kines may contribute to HPA-axis hyperactivity indirectly, as well,

through their effects on the glucocorticoid receptor. Cytokines in-

crease glucocorticoid receptor resistance by way of several signal-

ing pathways, including activation of the p38 MAPK, and by

stimulating changes in the expression of glucocorticoid receptor

isoforms (Pace et al., 2007). These changes, in turn, create a dysreg-

ulation of the CRH feedback system and result in a feed-forward

cascade that decreases the inhibitory effect of glucocorticoids on

CRH and stimulates increased cytokine production.

Further evidence on the neuromeodulatory effects of cytokines

comes from studies using functional magnetic resonance imaging

(fMRI) techniques. A number of studies have shown that inflamma-tory activation can stimulate changes in key areas of the brain that

underlie depression. For example, following experimental induc-

tion of an immune-inflammatory response, Eisenberger et al.

(2010) found that research participants showed increased symp-

toms of depression, and reduced activity in the ventral striatum

(VS)—the neural correlate of anhedonia—in response to reward

cues. In another study investigating the neural correlates of psy-

chomotor slowing, neural activity in the substantia nigra was

found to correspond to increased levels of IL-6 and decreased reac-

tion times following immune stimulation (Brydon et al., 2008). The

same research group has also reported findings that depressed

mood and increased circulating levels of IL-6 following immune

stimulation correspond to changes in areas of the brain highly in-

volved in emotional processing. Specifically, they found increasedsubgenual anterior cingulate cortex (sACC) activity and reduced

connectivity of sACC to mesolimbic regions of the brain including

the amygdala, medial prefrontal cortex, nucleus accumbens, and

superior temporal sulcus (Harrison et al., 2009).

4. The ‘‘infection-defense hypothesis’’ of depression

Advances in understanding the ways in which immune func-

tion, inflammation, and depression are related potentially provide

clues as to why genes that increase vulnerability to depression

have persisted at relatively high levels in the gene pool, despite

depression’s significant costs to reproduction and survival. The

infection-defense hypothesis of depression (Kinney and Tanaka,2009) offers an integrative view of these findings, based on the idea

that depression confers a critical compensatory advantage in im-

mune protection that has offset notable disadvantages of depres-

sion for evolutionary fitness. More specifically, the infection-

defense hypothesis proposes that immune vulnerability to infec-

tion elicits depressed mood, which in turn stimulates behaviors

that help protect vulnerable individuals and their kin against infec-

tious diseases (see Table 2).

The immune system is an amazingly sophisticated system

involving multiple, layered, and complementary strategies and

mechanisms for helping individuals combat infections, second only

in complexity to the nervous system. That natural selection has

sculpted this incredibly complex system is testament to what a

powerful role infection has played in natural selection. What we

are hypothesizing here is that natural selection may have also

sculpted coordinated pathways between the nervous and immune

systems so as to evolve behavioral response mechanisms that help

prevent and combat infections. Others, including Hart, have noted

the potential for such mechanisms: ‘‘It is quite logical to expect

animals and people to also have evolved nonimmunologic dis-

ease-fighting strategies including behavioral patterns, that might

serve as a first line of defense before the nonspecific and specific

immunologic systems are activated and that would complement

or potentiate immunological processes’’ (Hart, 1988 (p. 123)). As

Hart notes, behavioral responses, in principle, offer a valuable po-

tential advantage for combating infection—which is to avoid

becoming infected in the first place. There are numerous examples

of preventive immune strategies including, for example, human

aversion to noxious odors, which helps to prevent contact withhigh-pathogen sources such as human waste and decaying flesh.

Although at first glance it may seem counter-intuitive, from an

evolutionary perspective, it is precisely the high costs associated

with depression, combined with the fact that it is common and

heritable, that argues for some adaptive advantage to depression.

That is, given such negative consequences, there must be some

critical countervailing force or compensatory advantage that has

allowed susceptibility genes for depression to remain in the popu-

lation at high rates. Compensatory advantages in combating infec-

tion are a promising candidate, given that infection has had strong

associations with morbidity and mortality throughout evolution-

ary history. Increasing evidence that complex, deeply rooted mech-

anisms have evolved, by which the nervous and immune systems

can communicate and influence one another, also points in thisdirection.

Examples whereby costly adaptations have evolved because

they offer compensatory advantages in defending against infection

are numerous and widespread. Classic examples in humans in-

Table 2

Key postulates of the infection-defense hypothesis of depression.

1. Immune vulnerability to infectionelicits depressed mood, which in turnstimulates behaviors that help protect vulnerable individuals and their kin against infectious

diseases.

2. Depressive signs and symptoms offer individuals advantages for fighting existing infections by helping to conserve energy and avoid environmental stressors and

insults that could provide further immune-compromising challenge.

3. By inhibiting social contact, depression also prevents individuals and their biological relatives from contracting new infections.

4. Moods orchestrate—in a timely, titrated, and strategic manner—an array of behavioral and immunological responses to infections and immune vulnerability, which

can in turn be affected by a wide range of factors, including, e.g., genetic and seasonal variables, physical illness or injury, nutritional status, hormonal fluctuations,

exposure to environmental toxins, sleep disturbance, and stress.



5/14

clude sickle cell anemia and other hereditary hemoglobinopathies,

such as thallasemia. Sickle cell anemia, for example, is a hereditary

disease that is typically fatal, yet it is very prevalent in tropical re-

gions. People who inherit only one copy of the sickle-cell gene have

increased resistance to malaria, allowing the gene to persist de-

spite its severe costs (Kwiatkowski, 2005). An example from non-

human species involves the evolution of eye-catching plumage

and courtship dances that peacocks and the males of many other

avian species use to woo females. At first glance—from an evolu-

tionary perspective—this is a rather puzzling phenomenon; after

all, it takes significant energy and metabolic resources for the birds

to produce those ostentatious feathers and dances, and increases

visibility to predators. So why do it? Field studies indicate that it

is adaptive because showy feathers and skillful dance signal to

the females that the male carries fewer parasites. When the female

chooses the visually more attractive males, she reduces her own

exposure (and that of her offspring) to infection, and increases

the likelihood that the male (and her offspring) will have better ge-

netic resistance to infections in that ecological niche (see Kempe-

naers, 2007).

Behavioral strategies that defend kin against risk of infection

can also be noted in species such as honey bees and African naked

mole rats (see Preti, 2007). These species are like humans in that

they are social and live together in large, complex communities,

and have been known to display altruistic self-sacrifice, often leav-

ing the hive or burrow to die when they are ill—reducing the risk

that they will expose others within their community to the

infection.

From an evolutionary perspective, moods potentially provide an

attractive system for behavioral defense against infection, as they

have the ability to (1) orchestrate a wide range of behaviors; (2)

create responses that are generalizable to a variety of environmen-

tal challenges—a variety that reflects the inherent mutability and

variability of environmental pathogens; (3) respond to immune

vulnerability in a way that is both timely and titrated; and (4)

act preemptively, based on both endogenous and environmental

signals of increased infection risk. Depression’s overlapping fea-tures with sickness behavior appear to help conserve energy and

avoid further immune challenges, while depressive states are more

sustainable than sickness behavior over longer periods of time by

reducing the high cost of fever. This may allow depression to serve

as a complementary line of defense when the early immune re-

sponse to infection cannot be sustained during extended periods

of infection or immune vulnerability.

What we propose is even broader, however. A key aspect that

distinguishes the infection-defense hypothesis from many other

evolutionary theories of depression is that it proposes that depres-

sive signs and symptoms not only offer individuals advantages for

fighting existing infections, but also for preventing individuals and

their biological relatives from contracting new infections. Social

withdrawal, low energy, irritability, and blunted affect, for exam-ple, may greatly reduce the spread of infections by discouraging

mobility and/or social contact with others.

5. Some testable predictions of the infection-defense hypothesis

A number of testable predictions follow from the infection-de-

fense hypothesis. These predictions include the following:

1. Most signs and symptoms of depression will aid the immune

system’s ability to fight infections, by performing one or more

of the following functions:

(a) conserving metabolic resources for use by the immune sys-

tem in fighting infection;

(b) directly enhancing immune function through antimicrobialaction or by stimulating an increase in NK cell activity;

(c) reducing the risk of further environmental stresses that

impair immune function; and/or

(d) helping individuals and their kin to avoid transmitting or

contracting new infections.

2. Many types of infectious diseases will be associated with

depressive symptoms.

3. Depressed individuals will tend to have elevated rates of infec-

tion and/or immune alteration.

4. Medical, environmental, and physiological conditions that

increase immune vulnerability, or that increase exposure to

infection, will also be associated with increased rates of

depression.

5. There are bidirectional processes that communicate between

the nervous and immune systems and provide mechanisms

for infections, immune processes, and mood to influence one

another.

6. Moods provide an implicit mechanism for cost-benefit analysis

of an individual’s optimal responses to environmental chal-

lenges and the organisms’ immune status, helping to regulate

the timing and intensity of infection-defense responses.

Considerable evidence exists to support each of these predic-

tions. Evidence relevant to the respective predictions is discussed

in the following six sections.

6. Evidence for prediction # 1: most signs and symptoms of

depression appear to aid the immune system’s ability to fight

infections

As the infection-defense hypothesis predicts, there is evidence

that most features of depression appear to aid defense against

infections. Depressive signs and symptoms do this in four ways:

(a) conserving energy; (b) directly enhancing immune function

through antimicrobial action or by stimulating an increase in NK

cell activity; (c) reducing the risk of further immune-compromis-

ing challenges; and/or (d) by helping individuals and their kinavoid contracting new infections (see Fig. 1).

Akin to the symptoms of sickness behavior present during an

acute immune response to infection, depressive symptoms—such

as increased bodily aches and fatigue, hypersomnia, and psycho-

motor retardation—also reduce physical activity and encourage

rest, thereby conserving metabolic resources for use by the im-

mune system. In addition to these somatic and behavioral changes

that occur during depression, anhedonia and cognitive features of

depression—such as decreased self-esteem and feelings of hope-

lessness (American Psychiatric Association, 2000)—also discourage

exploration and risk-taking, thereby reducing exposure to new

pathogens or to accidents, conflicts, or injuries that can exacerbate

existing immune system compromise. A number of studies have

found, for example, that depressed individuals exhibit state-dependent increases in harm-avoidance behavior (e.g., de Winter

et al., 2007; Nery et al., 2009). Moreover, the degree of harm-avoid-

ance behavior endorsed by depressed individuals increases along

with increases in depression severity and the number of depressive

episodes (Nery et al., 2009).

Moods help to regulate social behavior. Social withdrawal and

decreased sexual drive, as well as cues that discourage social con-

tact from others—such as decreased speech, irritability, and

changes in body language and facial expression (blunted affect)—

can reduce the risk for contracting or transmitting infections,

including sexually transmitted ones. Thus, the changes in social

interest and demeanor associated with depression not only reduce

an individual’s risk for contracting new infections by discouraging

close contact with others, but also reduce the risk of transmittinginfection to relatives. The importance of reducing the spread of



6/14

infection to others is highlighted by many examples throughout

human history whereby infectious diseases were known to wipe

out entire families or villages (e.g., Thwaites et al., 1997). The nas-

cent immune system of infants and young children render them

especially vulnerable to infections. Consistent with the notion of

a heightened need to protect the young, depression among human

mothers, as well as sickness behavior among murine animal mod-

els, is associated with disengagement—and sometimes even rejec-tion—of offspring (Burke, 2003; Dantzer, 2009).

Reduced mobility and reduced sociability associated with

depression, as noted by Raison and Miller (2012), may also have

the effect of minimizing exposure to out-group members. Contact

with individuals from different groups or environments who are

immunologically dissimilar increases the risk of exposure to patho-

gens to which an individual has not developed specific immunity

in their home environment.

Appetite changes are extremely common in depression. De-

pressed individuals can vary in whether they tend to increase or

decrease overall intake of food (American Psychiatric Association,

2000). However, the specific forms that these appetite changes

take in depression may constitute alternative strategies that help

defend against infection. Proinflammatory cytokines, for example,have a number of known anorectic effects, including increased

stimulation of the peptide leptin, which plays a key regulatory role

in balancing food intake and energy expenditure (Andréasson et al.,

2007; Wong and Pinkney, 2004). Reduced appetite can reduce the

risk of food–borne parasites, a major source of infection through-

out history and the present (Mead et al., 1999). By contrast, some

forms of depression, including seasonal affective disorder (SAD),

are associated with increased carbohydrate cravings. Carbohy-

drates and carbohydrate-based foods, such as sugar and bakedbread, offer high metabolic and caloric resources, while at the same

time being less likely to cause illness due to spoiling than many

other foods, such as meat, fish, or fruits and vegetables. In addition,

increased intake of carbohydrates influences immune function by

triggering changes in proinflammatory cytokines and an increase

in NK cell activity (see Braun and Von Duvillard, 2004). It has also

been noted that high intake of carbohydrates can promote defense

against infection by decreasing the relative intake of lipids, as lipid

consumption has been associated with worse outcomes for infec-

tious illness (Heyland et al., 1998).

Some individuals with depression develop paradoxical symp-

toms that include notable anxiety, agitation, and insomnia, consti-

tuting a state of hypervigilance. In spite of high metabolic costs,

heightened vigilance is likely to have been an extremely adaptiveresponse to stress throughout human evolution, where the threat

Fig. 1. A neuro-immune feedback system regulates the timing and intensity of behaviors that aid defense against infection. Immune vulnerability is influenced by acute risk

of infection, as well as by a number of other known immune-compromising factors that include stress, sleep deprivation, chronic infection and disease, winter season,

hormonal fluctuations, exposure to toxins, and nutritional deficiencies. According to the infection-defense hypothesis, depression and related behaviors that aid against

infection are activated by immune status, which in turn reduce immune vulnerability and risk of infection by (a) conserving or enhancing metabolic resources for fightinginfection, (b) directly aiding immune defenses through antimicrobial action and/or by increasing NK cell activity, (c) reducing risk of further immune challenge, such as

avoiding danger or conflict to reduce the risk of physical injury, or restricting patterns of eating to avoid food-borne illness, and/or (d) by reducing the spread of infection

through decreased social contact. 1 Depressive symptoms vary by subtype, accounting for seemingly paradoxical symptoms involving energy and appetite. For example,

seasonal affective disorder (SAD) and other atypical forms of depressionoften include symptoms of psychomotor slowing, hypersomnia, and increased appetite/carbohydrate

cravings, while other forms of depression – such as melancholia – include symptoms of restless agitation, anorexia, and insomnia.



7/14

of predators and other physical dangers to self and family was high

(Marks and Nesse, 1994). Hypervigilance may contribute to harm-

avoidance behaviors that, as noted earlier, reduce risk of exposure

to new infections or immune-compromising injuries and stressors.

Furthermore, although speculative, restless pacing or hand-wring-

ing, which can occur during agitated forms of depression, may offer

some benefits to immune function just as other forms of moderate

physical activity do. For example, moderate physical activity can

increase NK cell activity (Nieman et al., 2005) and elevate core

body temperature, which may have additional immune effects

(Mestre-Alfaro et al., 2012). An agitated state tends to occur in

the context of patients’ increased anxiety about their personal

safety and that of their family, thereby discouraging them from

adventurous activity outside the home, so that the increased exer-

cise occurs in a way that reduces the risk of injury and contact with

new infections that would otherwise tend to occur as a result of in-

creased exercise.

7. Evidence for prediction # 2: many types of infection are

associated with depression

Increased rates of depressive symptoms have been observed

following an individual’s contracting a number of acute or suba-

cute infections (Fazekas et al., 2006; Murray et al., 2007), as well

as following chronic infections (Cohen et al., 2002; Lipkin and Hor-

nig, 2004; O’Connor et al., 2009; Yates and Gleason, 1998), and

immunization with live virus vaccines (Afsar et al., 2009; Glaser

et al., 2003; Yirmiya et al., 2000). For example, increased depres-

sive symptoms have been found in patients with herpes simplex

encephalitis (Fazekas et al., 2006) and West Nile virus (WNV)

(Murray et al., 2007). Murray et al. (2007) found that over 31% of

WNV patients in their study experienced the onset of a depressive

episode within one year of their becoming infected. Those who had

been diagnosed with the more severe form of WNV, neuroinvasive

disease, showed an even greater risk for developing depression.

Evidence for an increased risk of depression among those whosuffer from chronic infections comes from a variety of studies

involving both humans and laboratory animals. Depression-like

behavior has been observed, for example, in mice infected with Ba-

cille Calmette Guerin (BCG), a bacterium related to the one that

causes tuberculosis (O’Connor et al., 2009). Humans suffering from

chronic infections, such as human immunodeficiency virus (HIV),

hepatitis C (HCV), and genital herpes (HSV-2), have also been found

to have an increased risk for depression (Yates and Gleason, 1998;

Lipkin and Hornig, 2004). Among HIV patients, high viral loads are

associated with higher depression scores on the Hamilton Rating

Scale for Depression (Cohen et al., 2002).

Vaccines containing live viruses can also cause depressive

symptoms. For example, in a prospective double-blind study (Yir-

miya et al., 2000), vaccination with live attenuated rubella virus re-sulted in depressive symptoms that lasted up to 10 weeks. In

another study, Afsar et al. (2009) found increases in symptoms of

depression related to antibody response following Hepatitis B vac-

cination in hemodialysis patients. Among certain sub-groups of

older adults, influenza vaccination has also been linked to in-

creased depressive symptoms, accompanied by a rise in serum lev-

els of IL-6 (Glaser et al., 2003).

8. Evidence for prediction # 3: elevated rates of infection and/or

immune alteration are found in depressed patients

Just as elevated rates of infection would be expected among

individuals with fever, elevated rates of infection and/or immunealteration are expected in depressed patients, given our proposal

that depression is elicited as an immune defense during conditions

of infection or immune-compromise.

Accordingly, depressed individuals show marked evidence of

immune suppression and excessive inflammation (Dowlati et al.,

2010; Irwin and Miller, 2007; Zorilla et al., 2001), as well as greater

vulnerability to infection (Cohen, 1995; Irwin, 2002b; Zorilla et al.,

1996). For example, a robust association has been found between

depressive symptoms and reduced NK cell activity (Zorilla et al.,

2001; Reiche et al., 2004). Depression has also been associated

with decreased lymphocyte proliferation, although the results are

less consistent, and appear to depend on moderating factors such

as age (Schleifer et al., 1989).

Increased levels of proinflammatory cytokines and chronic

inflammatory response have been found repeatedly in depressed

individuals (Dowlati et al., 2010; Irwin and Miller, 2007; Zorilla

et al., 2001). Depressed individuals most often show evidence for

increased levels of the proinflammatory cytokines IL-1b, IL-6,

TNF-a, and INF-c (Maes et al., 2009). Other inflammatory biomark-ers that have shown regular associations with depressive symp-

toms include changes in acute-phase-response protein, C-reactive

protein (CRP), haptoglobin, nitric oxide, and glucocorticoids (Loftis

et al., 2010; Maes et al., 2009; Zorilla et al., 2001 ).

A number of studies indicate that depressed individuals show

increased vulnerability to contracting infections, such as upper

respiratory tract infections (Cohen, 1995). Among HIV-infected pa-

tients, depressive symptoms are associated with higher viral load

and lower NK cell activity (see review by Kopinsky et al., 2004).

Depressive symptoms are also significantly associated with recur-

rence of herpes simplex infection (Zorilla et al., 1996) and reduced

cellular immunity to varicella-zoster (Irwin, 2002b).

9. Evidence for prediction # 4: many conditions associated with

immune vulnerability are also associated with depression

The list of conditions that show associations both with in-

creased vulnerability to infection and with elevated risk for depres-sion is long and varied, and includes seasonal factors, nutritional

deficiencies, hormonal fluctuations, toxin exposure, cancer, cardio-

vascular disease, autoimmune diseases, and chronic pain (see Ta-

naka et al., 2012, for a review). Two widely-studied factors linked

both to immune vulnerability and to depression include stress

and sleep deprivation. Stress, for example, which has a clear and

well-established role in the etiology of depression (Brown et al.,

1986; Hammen, 2005; Kendler et al., 1999; Sillaber et al., 2009),

is also associated with increased numbers of circulating proinflam-

matory cytokines (Bierhaus et al., 2003; Wolf et al., 2009). In-

creased cytokines sensitize the HPA-axis, contributing to the

hypersecretion of cortisol (Turnbull and Rivier, 1995), which has

been identified as a critical pathway in central nervous system

(CNS) dysregulation. Cortisol can suppress cellular immune re-sponse that is critical in defending the body against viral infections

(Maes et al., 1994). Accordingly, there is evidence that chronic

stress significantly increases vulnerability to infection (e.g., Cohen,

1995; Cohen et al., 1999; Horesh et al., 2008; Kiecolt-Glaser et al.,

2002; Mundt et al., 2000). For example, it has been found that

stressed caretakers are more vulnerable to infectious diseases (Kie-

colt-Glaser et al., 2002), and populations that experience higher

levels of social stress tend to have higher rates of mortality and

morbidity from infectious disease (Weiss and McMichael, 2004).

Another major risk factor for depression is chronic insomnia

(Lustberg and Reynolds, 2000). Chronic sleep deprivation is associ-

ated with increased immune vulnerability (Irwin, 2002a), and the

severity of insomnia is negatively correlated with NK cell activity

in both depressed and non-depressed groups (Zorilla et al., 2001;Irwin, 2002a). Poor sleep quality and insomnia can precipitate



8/14

inflammatory processes, raising plasma cortisol, C-reactive protein

(CRP), and proinflammatory cytokines (Motivala et al., 2005). In

experiments with rats, chronic restriction of sleep causes depres-

sion-like changes in both the sensitivity of neurotransmitter recep-

tors and neuroendocrine reactivity to stress (Novati et al., 2008).

In addition to stress and sleep deprivation, the list of conditions

that are associated both with immune vulnerability and with

depression is staggering and diverse. They include seasonal factors

such as reduced daylight hours and colder temperatures during the

winter months (Lam et al., 2004; Magnusson, 2000), nutritional

factors such as deficiencies of Vitamin D or Omega-3 fatty acids

(Borges et al., 2011; Ganji et al., 2010; Hibbeln, 2009; McNamara

et al., 2010; Parker et al., 2006), hormonal fluctuations associated

with the menstrual cycle and post-partum period (Groër and Mor-

gan, 2007; Rubinow et al., 2009), exposure to environmental toxins

such as toxigenic mold and pesticides (Anyanwu et al., 2003; Bes-

eler and Stallones, 2008; Corsini et al., 2008; Udoji et al., 2010), and

a number of medical or physiological conditions such as cancer

(Miller et al., 2008), cardiovascular disease (Cesari et al., 2003;

Kling et al., 2007), autoimmune diseases (Gold and Irwin, 2006;

Nery et al., 2008), and chronic pain (Watkins and Maier, 2000).

That such a wide array of conditions has well-established links

with both depression and immune alteration is consistent with

the notion that depression evolved as a generalized immune re-

sponse that can offer broad defense against a range of immune

challenges. A preponderance of cross-sectional research in these

areas limits causal interpretations and highlights the need for more

longitudinal investigations; however, a number of experimental

studies have reliably demonstrated that conditions associated with

sources of immune vulnerability such as stress and sleep depriva-

tion increase risk of depression-like behavior in animal models

(e.g., Ardayfio and Kim, 2006; Novati et al., 2008; Sillaber et al.,

2009).

10. Evidence for prediction # 5: bidirectional links between the

nervous and immune systems allow moods and immune vulnerability to influence each other

Multiple pathways have been identified by which immune sys-

tem activation may lead to depression, including activation of the

HPA-axis by proinflammatory cytokines (Holsboer, 2000; Pace

et al., 2007) and degradation of tryptophan by activation of the

IDO (Capuron et al., 2003; Dantzer et al., 2008). There is now

well-established evidence to suggest that bidirectional communi-

cation between the immune system and nervous system also oc-

curs (Maier and Watkins, 1998).

Bidirectional communication between the immune and nervous

systems is suggested by the immune-enhancing effects of anti-

depressant interventions, such as electroconvulsive therapy (ECT)

(Fischler et al., 1992; Kronfol et al., 2002; Hestad et al., 2003), avariety of psychosocial interventions (Fang et al., 2010; Robinson

et al., 2003; Witek-Janusek et al., 2008), and meditation practice

(Lavretsky et al., 2012; Pace et al., 2009). For example, while

ECT—known for its rapid amelioration of depressive symptoms—

shows no evidence of changes in monoamine metabolism or neu-

roplasticity following a single ECT session, augmented NK cell

activity is found after a single session (Fischler et al., 1992; Kronfol

et al., 2002), with a gradual and significant decline in TNF-a levelsover the course of repeated sessions (Hestad et al., 2003).

There is evidence that NK cell activity is significantly elevated

after experimental exposure to pleasurable conditions (Berk

et al., 2001; Matsunaga et al., 2008), and that psychological inter-

ventions for the treatment of depression and emotional distress,

such as mindfulness training and cognitive behavior therapy(CBT), also enhance NK cell activity (Masuda et al., 2002; Robinson

et al., 2003; Witek-Janusek et al., 2008). Notably, among partici-

pants in a mindfulness-based stress reduction program, Fang

et al. (2010) found that NK cell activity was significantly enhanced

only in patients who reported improved well-being following the

intervention.

Studies of meditation practice also show notable effects on

stress and immune responses. For example, Pace et al. (2009)

found that among individuals undergoing training for compassion

meditation, the amount of time spent in meditation practice was

associated with decreased IL-6 concentrations and decreased levels

of distress following exposure to a laboratory stressor. In another

study (Lavretsky et al., 2012), dementia caregivers who practiced

daily yogic meditation experienced a reduction in depressive

symptoms and reduced signs of stress-induced aging, measured

by telomerase activity.

Several types of antidepressants and mood stabilizers are also

noted to have antibiotic effects. In a recent review, Lieb (2007)

cites a number of in vivo and in vitro studies demonstrating that

antidepressant drugs and lithium have significant antimicrobial

and immune-potentiating effects. These effects include the ability

to reverse resistance of bacteria to antibiotics (e.g., Kristiansen

et al., 2010). It has been proposed that psychotropic medications

derive their antimicrobial properties by acting as bacterial efflux

pump inhibitors (EPIs) (Munoz-Bellido et al., 2000). Lieb (2007)

also notes that several antibiotics have been found to have

mood-enhancing effects.

11. Evidence for prediction # 6: moods orchestrate the timing

and intensity of infection-defense behaviors

There is evidence that moods also provide a mechanism for reg-

ulating the timing and intensity of immune-related behaviors

based on an implicit cost-benefit analysis of what is most adaptive

for survival and reproduction. Depressive behaviors are biologi-

cally costly, interfering with work and social functioning, and espe-

cially in pre-modern times would likely interfere with efforts toacquire and compete for food, territory, and mates. Therefore, the

benefits would outweigh the costs only when initial efforts to cope

with an immune challenge are ineffective, or when an individual

suffers from chronic infection or stress. Sickness behavior and

depression (and perhaps various sub-types of depression) are elic-

ited—and vary in timing and intensity—based on feedback from the

immune system. An extension of this is that mood elevation can

stimulate more adventurous or gregarious behavior when an indi-

vidual is in a more robust immune state.

Evidence from a number of studies indicates that risk for

depression and depression severity varies in relation to immune-

compromise. For example, as mentioned above, Murray et al.

(2007) found a heightened risk for depression among a group of

individuals who had contracted WNV disease, while those whohad been diagnosed with the more severe form of WNV, neuroin-

vasive disease, demonstrated an even greater risk for depression.

In a study of HIV patients, high viral loads were associated with

higher depression scores on the Hamilton Rating Scale for Depres-

sion (Cohen et al., 2002). Maes et al. (1994, 1995) have also found

depression severity to correlate with circulating numbers of proin-

flammatory cytokines, including IL-1 and IL-6, as well as with mea-

sures of HPA-axis hyperactivity.

There is also evidence that the timing of mood response to anti-

depressant interventions parallels changes in immune function

(Lieb, 2007; Tanaka and Kinney, 2011b). In a recent paper, Tanaka

and Kinney (2011b) reviewed studies that provided information

about the time course of mood response and NK cell activity fol-

lowing a number of mood-enhancing interventions such as exer-cise, therapeutic administration of ketamine, and antidepressant



9/14

treatments. Notable parallels in the time course of mood improve-

ment and immune response were observed across all modalities.

For example, both mood and NKCA are rapidly enhanced in re-

sponse to exercise, and both of these effects tend to dissipate with-

in several hours after the conclusion of exercise, following a similar

time course. In a separate review, Lieb (2007) described evidence

that antidepressant drugs and lithium demonstrate concomitant

increases in mood-response and immune-potentiating effects.

12. Comparison of the infection-defense hypothesis with other

evolutionary theories of depression

Although the proliferation of recent evidence linking depression

to inflammation and immune function has led some to speculate

about the potentially adaptive role of depression in helping fight

infections, there is currently little in the way of systematic at-

tempts to help explain depression from an evolutionary infec-

tion-defense point of view. An exception to this is Raison and

Miller’s (2012) recent hypothesis of pathogen host defense

(PATHOS-D). Drawing on evidence that many genes identified as

risk alleles for depression are also associated with immune factors

related to pathogen host defense—such as hyperthermia, reduced

bodily iron stores, conservation/withdrawal behavior, hypervigi-

lance, and anorexia—PATHOS-D, like the infection-defense hypoth-

esis, suggests that the advantages in pathogen host defense

conferred by depression risk alleles offset depression’s notable

costs to reproduction and survival. The two hypotheses differ,

however, in their explanations of how depression relates to im-

mune activation. According to PATHOS-D, genes that are involved

in immune response and depression are one in the same, and

depression is intrinsically tied to immune activation. The infec-

tion-defense hypothesis, on the other hand, can be theoretically

de-coupled from immune activation, suggesting that well-estab-

lished mechanisms for communication between the immune sys-

tem and the nervous system provide pathways for depression to

be triggered and modulated by infections as well as other signalsof vulnerability, which is consistent with evidence that depression

is commonly—but not always—associated with an immune-inflam-

matory response (Raison and Miller, 2011).

More traditional attempts to explain the persistence of depres-

sion from an evolutionary standpoint have focused on socially

adaptive mechanisms. Evolutionary theorists have often drawn

upon key features of depression—such as sadness and with-

drawal—as part of a subversive, ‘‘second-line’’ posture that offers

protective advantages in response to defeat, and helps to explain

many of depression’s puzzling, self-limiting features. For example,

according to ‘rank theory,’ depression signals yielding behavior

following an unsuccessful struggle for dominance in order to

avoid unnecessary harm (Gilbert, 1992). Similarly, ‘social risk the-

ory’ describes depression as a signal of submissiveness in re-sponse to social failure, as a way to maintain social ties critical

for survival and reproduction (Allen and Badcock, 2006; Price

et al., 1994). The ‘psychic pain hypothesis,’ offers a more general

model of depressed mood, suggesting that it provides affective

feedback that discourages continued investment in adverse or

unreachable goals (Thornhill and Thornhill, 1989). Other theories

describe ways in which behavioral changes and nonverbal cues

associated with depression function as distress signals to elicit

help during adversity (Hagen, 2003). Each theory hones in on

selection pressures that have likely shaped affective responses

throughout evolution; sadness, for example, heightening attention

to adversity, provides important feedback to self and others about

social and environmental sources of stress, and can promote adap-

tive behavior in response to the types of losses or conflicts de-scribed above. From our infection-defense perspective, the

immune system co-opts this affective response so that the deep,

pervasive sense of sadness that often occurs in the context of clin-

ical depression may help to promote withdrawal behaviors that

conserve energy and reduce exposure to new sources of stress

or infection.

Traditional evolutionary theories of depression, while offering

potentially useful explanations for understanding changes in af-

fect and behavior under a circumscribed set of conditions, on

the whole remain limited (1) in their ability to offer predictions

that can be submitted to rigorous testing; (2) in their ability to ex-

plain the full range of signs and symptoms of depression, includ-

ing apparently paradoxical features such as hypervigilance vs.

conservation/withdrawal states; (3) in their ability to explain

depression’s association with a diverse array of conditions and ill-

nesses such as nutritional deficiencies, seasonal changes, hor-

monal fluctuations, and chronic diseases; and (4) in their failure

to account for evidence that depression is linked to immune-

inflammatory factors, mediated by a variety of neuroimmune pro-

cesses. The infection-defense hypothesis, by contrast, helps to

integrate a large and growing body of research that depression

is intimately linked with immune function, and offers insight into

the mechanisms by which many known risk factors for depres-

sion, such as stress and sleep deprivation, contribute to the etiol-

ogy of depression.

13. Potential challenges to the infection-defense hypothesis

Depression’s associations with higher morbidity and mortality

rates for a number of diseases appear to contradict the notion that

depression provides an evolutionary advantage. From the infec-

tion-defense point of view, however, we would expect depression

to be elicited more often in the context of illness and disease; thus,

there would be extremely high rates of morbidity and mortality

associated with it. There is some evidence to suggest that depres-

sion not only follows, but also presages worse outcomes for a num-

ber of illnesses, including HIV (Leserman, 2008) and cancer (Reicheet al. 2004). More longitudinal data is needed to provide an accu-

rate account of the progression and relative contributions of

depression, inflammation, and potential ‘‘third variables’’ that

may be influencing immune response. It is important to note, how-

ever, that our hypothesis only requires that the relative advantages

of depression outweigh its costs. Throughout human history, the

impact of infection on survival and reproduction has been stagger-

ing, and life expectancy throughout most of human history was

less than half of what it is today; thus, immune responses are likely

to be biased toward surviving acute or relatively short-term im-

mune challenges, while the long-term effects of chronic immune-

activation or inflammation only become more relevant as life

expectancy gradually increases over time. Moreover, from a Dar-

winian perspective, the fact that depression has persisted despitethese costs only strengthens the case that it must have offered a

powerful adaptive advantage for risk alleles to have remained so

high in the population.

There may be other ways to understand depression’s persis-

tence from an immune standpoint. For example, might depression

simply be an epiphenomenon of a prolonged or severe inflamma-

tory response? Or is it possible that depression is a maladaptive

expression of genes that confer immune advantages to the host un-

der some circumstances, and confer risk in others? Just as fever, a

key asset in combatting infections, can be harmful or even lethal at

high levels, an adaptive view of depression does not imply that

depression can’t sometimes be harmful. However, the view that

depression is a fundamentally maladaptive state is again refuted

by the notably high prevalence rates of depression risk alleles inthe population.



10/14

14. A comprehensive view of depression

A comprehensive view of depression takes into account its mul-

ti-faceted nature that manifests in a variety of clinical presenta-

tions. For example, as noted above, some depressed individuals

experience hypervigilant states that include agitation and restless-

ness, while others experience low energy and psychomotor slow-

ing (American Psychiatric Association, 2000). At any given pointin time, the specific features of depression that are evoked likely

depend on the interaction of genetic factors (Raison and Miller,

2012), early experience (Bradley et al., 2008; Chapman et al.,

2004; Garnefski et al., 1990), immune vulnerability (Dantzer

et al., 2008; Kinney and Tanaka, 2009), physiological vulnerabili-

ties (Maes et al., 2009; Raison and Miller, 2011), and situational

triggers (Keller and Nesse, 2006). For example, grief-related

depression following the loss of a loved one is more likely to elicit

crying, whereas depression that follows from stress or seasonal

factors is more likely to involve pessimism and fatigue (Keller

and Nesse, 2006). While not all individuals who experience depres-

sion show evidence of increased inflammation (Raison and Miller,

2011), it remains to be seen whether some forms of depression are

better explained from alternate bio-behavioral frameworks. For

example, grief reactions to separation and loss, rooted in survival

strategies employed during the vulnerable state of infancy, may

be better explained from an ethological attachment framework

(Bowlby, 1982; Mikulincer and Shaver, 2007).

Gene-environment interactions have been found to exert influ-

ence on depression risk, as in the case of the 5-HTTLPR short allele

of the serotonin transporter gene. Numerous studies have shown,

for example, that carriers of the 5-HTTLPR short allele demonstrate

an elevated risk for depression that is further increased in relation

to levels of environmental stress (Eley et al., 2004; Kendler et al.,

2005).

Evidence from studies using both human and laboratory ani-

mals suggest that early life experiences can influence susceptibility

to inflammation and depression. Early adverse experiences such as

neglect and trauma are related to increased risk for depression anddisease across the lifespan (Chapman et al., 2004; Garnefski et al.,

1990). Childhood maltreatment is also associated with increased

inflammation (Danese et al., 2007), and adults with a history of

childhood maltreatment have been found to respond to acute

stress with increased IL-6 concentrations compared to controls

(Carpenter et al., 2010). Laboratory studies using rats have found

that impaired maternal contact and milk quality can induce signif-

icant changes in stress response and inflammation in offspring

(Walker, 2010). Increased CRH concentrations and HPA-axis activ-

ity are found in both humans and laboratory animals who have

experienced early life stress, and have been implicated as key fac-

tors mediating the associations between early experience, inflam-

mation, and depression (Bradley et al., 2008; Gillespie et al.,

2009; Walker, 2010).Given the complex interplay of multiple systems involved in a

depressive response, immune response may also interact with

other physiological vulnerabilities to increase risk for depression.

Maes and colleagues (Maes et al., 2009), for example, have outlined

factors related to the condition of increased gut permeability,

known as ‘‘leaky gut,’’ that may contribute to inflammation-associ-

ated depression, such as increased translocation of lipopolysaccha-

ride (LPS) from gram-bacteria.

While the infection-defense hypothesis posits a protective role

for depression in humans’ environment of evolutionary adapted-

ness, the role of increased hygiene and availability of antimicrobial

treatments in modern society have reduced mortality rates associ-

ated with infection, and suggests the possibility that depression’s

relative advantages in infection-defense may be decreasing. Thefact that rates of depression are increasing does not necessarily

counter this, as the increasing rates may not be due to increased

genetic risk per se, but rather to the increased presence of im-

mune-compromising factors such as environmental toxins.

It is also of note that, from a clinical perspective, psychological

interpretations of depression—including psychodynamic or sche-

ma-based interpretations—insofar as they help to identify and alle-

viate sources of psychological stress, including classically-

conditioned stress responses to situational triggers, complement

an infection-defense view of depression. The infection-defense

view of depression not only helps to provide an evolutionary

explanation for the prevalence and persistence of depression

throughout human history, but also supports an integrative frame-

work for understanding the etiology of depression from multiple

levels of influence.

15. Summary and further clinical implications of the infection-

defense hypothesis

Converging evidence suggests that depression is often an

inflammatory/immune-mediated response to infection, vulnerabil-

ity to infection, and/or chronic activation of the innate immune

system. This inflammatory response is stimulated by increasedproduction of proinflammatory cytokines, which have wide-rang-

ing effects on both neuroendocrine and neuronal systems, includ-

ing an inhibitory influence on serotonergic transmission. This

emergent model of depression helps to explain: (1) why depres-

sion is associated with immune alterations such as decreased

NKCA, (2) why depression is associated with increased rates of

infection and disease, and (3) why a wide range of environmental

and physiological factors associated with increased vulnerability to

infection are also associated with increased risk for depression.

In addition, the discovery of inflammatory-immune factors in

the physiology of depression helps to explain an important psychi-

atric puzzle as to why genes associated with major depression

have persisted, despite depression’s association with increased

morbidity and mortality. As outlined in the infection-defensehypothesis, signs and symptoms of depression such as anhedonia,

social withdrawal, reduced energy and psychomotor retardation—

and the genes that contribute to them—may be explained as adap-

tive responses to infection vulnerability that serve to: (1) conserve

metabolic resources for fighting infection, (2) reduce exposure to

further infections or environmental stressors, and (3) reduce social

contact to prevent the spread of infection to kin.

Unlike many other evolutionary theories of depression, key te-

nets derived from the infection-defense hypothesis can be submit-

ted to rigorous testing. For example, using a double-blind placebo-

controlled experimental design, a stringent test of the hypothesis

that interventions reducing infection and/or immune compromise

will lead to a reduction in depressive symptoms could be made. Fu-

ture studies using longitudinal and epidemiological data will also

be important in helping to establish the relative progression of im-

mune-inflammatory factors and depression following exposure to

infection or immune-compromising conditions.

15.1. Clinical Implications

Associations between immune function and depression, and in

particular, the notion that moods may serve as a behavioral de-

fense against infection, carry important implications for under-

standing the causes, treatment, and prevention of depression.

Some of these implications pose a challenge to conventional wis-

dom; for example, if symptoms of depression are serving a protec-

tive function in the face of immune challenge, then the goal of

reducing depressive symptoms to ease emotional distress shouldbe balanced against the need to first identify and treat possible



11/14

underlying immune factors. Some factors that influence immune

vulnerability, such as stress and sleep disturbances, are commonly

addressed in mental health treatment settings and can be allevi-

ated with traditional forms of therapy including CBT and mind–

body relaxation techniques, as well as pharmacological interven-

tions; however, the potential role of underlying infections and im-

mune disorders warrants greater attention. Underlying infections

or immune disorders may be under-diagnosed, and should rou-

tinely be ruled out as potential causative factors. This need is

underscored by the findings of Rabinowitz et al. (1997) that psy-

chiatric patients tend to have higher rates of unrecognized and un-

treated physical illness than the rest of the population.

The notion that depressed individuals may have increased im-

mune vulnerability, an underlying infection, or other source of im-

mune compromise leads to several further implications. First,

increased attention to hygiene in clinical settings such as waiting

rooms, shared bedrooms on inpatient units, or other shared spaces

may help to reduce the spread of pathogens that contribute to

depression. Relatedly, depressed individuals may benefit from tak-

ing heightened precautionary measures to reduce the risk of incur-

ring further immune challenge, such as more thorough and

frequent handwashing, or by avoiding crowded public places

where there is greater exposure to pathogens. Finally, engaging

in strategies to enhance immune function could potentially com-

plement traditional anti-depressant therapies. The reduction of im-

mune-compromising factors—such as sleep disorders, chronic pain,

stress, dietary deficiencies, and insufficient exercise—deserves

more investigation as a possible approach to treating and prevent-

ing depression. For example, moderate exercise, meditation prac-

tice, and nutritional supplementation to correct for deficiencies

in omega-3 fatty acids and vitamin D, may be simple and econom-

ical ways to help alleviate depressive symptoms.

Environmental pathogens that trigger immune impairment and

inflammation, such as toxigenic molds, pesticides, and other pollu-

tants, also deserve greater attention for their potential role in the

etiology of depression. For example, exposure to toxigenic mold

is associated both with alterations in both NK cell activity and var-ious neuropsychiatric symptoms, including depression (Anyanwu

et al., 2003). Increasing levels of environmental toxins that contrib-

ute to inflammation and depression may potentially help to ex-

plain recent increases in rates of depression (Compton et al., 2006).

Pharmacologic interventions that have established antimicro-

bial, immune-enhancing, or anti-inflammatory properties also

merit more study for their potential antidepressant effects. For

example, minocycline, a second-generation tetracycline antibiotic,

has demonstrated anti-inflammatory and anti-depressant effects

in both human and animal studies (Pae et al., 2008). Preliminary

findings suggest that non-steroidal anti-inflammatory drugs (NSA-

IDs) such as aspirin, ibuprofen, celecoxib, and naproxen may have

antidepressant effects by way of their action as cyclooxygenase

(COX)-2 inhibitors (Müller, 2010). For example, celecoxib has beenfound to block COX-2 enzymes as well as to reduce depression-like

behaviors in laboratory animals (Guo et al., 2009). In at least one

human study, celecoxib has been found to augment the anti-

depressant effects of fluoxetine (Akhondzadeh et al., 2009). This

latter finding, in particular, points to the possibility that COX-2

inhibitors may be effective adjunct therapies to traditional

antidepressants.

Another promising line of research in the search for improved

pharmacologic treatments of depression involves anti-inflamma-

tory cytokine antagonists. For example, in animal studies, the use

of the IL-1b receptor antagonist IL-1ra prevented the development

of depression-like behavioral and neurochemical changes in re-

sponse to chronic stress exposure (Koo and Duman, 2008; Norman

et al., 2010). The anti-inflammatory cytokine IL-10 has also beenidentified as a potential target for antidepressant action. In studies

with laboratory animals there is evidence that IL-10 modulates

numerous features associated with depression including sleep,

helplessness, and pain perception (Roque et al., 2009). Moreover,

increased levels of circulating IL-10 have been found in both hu-

mans and animals receiving anti-depressant treatment (Kenis

and Maes, 2002).

TNF-a inhibitors have shown promise in clinical trials fordecreasing symptoms of depression. Patients with treatment-resis-

tant depression who had high baseline levels of inflammatory bio-

markers showed improvement following several infusions of the

TNF antagonist Infliximab over a 12-week trial (Raison et al.,

2012). Another TNF inhibitor, Etanercept, administered to psoriasis

patients over a 12-week period, resulted in a greater decrease of

depressive symptoms from baseline when compared with placebo

(Tyring et al., 2006).

Of note, a potential risk of using anti-inflammatory agents in

the treatment of depression is that the suppression of immune

function may mask existing infections or increase susceptibility

to new ones. This again underscores the need to rule out infection

or immune-related causes of depression prior to treatment, and to

weigh the potential risks of anti-inflammatory/immunosuppres-

sant agents against the need to alleviate depressive symptoms.

In summary, the infection-defense hypothesis bears important

clinical implications for the treatment and prevention of depres-

sion. Its basic premise that depression is an adaptive immune

strategy (1) indicates that many behavioral features of depression

may serve an adaptive purpose to help fight existing infections and

avoid new ones; and (2) points to the need for a fundamental shift

in depression treatment that favors the primacy of investigating

and remedying underlying infections or immune-compromising

factors (both environmental and endogenous) that may be contrib-

uting to depression. In addition, interventions that aid the body’s

natural ability to fight infections, such as correcting nutritional

deficiencies or engaging in stress-reduction activities like medita-

tion, would be expected, as numerous studies show, to help temper

symptoms of depression. Finally, the proliferation of new discover-

ies over the past several decades that link depression to immune-inflammatory processes, and that have identified bidirectional

pathways by which the nervous and immune systems communi-

cate, also opens the door for a wide range of novel pharmacologic

interventions in the clinical management of depression.

Conflict of interest

All authors declare that there are no conflicts of interest.

References

Afsar, B., Elsurer, R., Eyileten, T., Yilmaz, M.I., Caglar, K., 2009. Antibody response

following hepatitis B vaccination in dialysis patients: does depression and

quality of life matter? Vaccine 27, 5865–5869.Akhondzadeh, S., Jafari, S., Raisi, F., Nasehi, A.A., Ghoreishi, A., Salehi, B., Mohebbi-

Rasa, S., Raznahan, M., Kamalipour, A., 2009. Clinical trial of adjunctive

celecoxib treatment in patients with major depression: a double blind and

placebo controlled trial. Depress. Anxiety 26, 607–611.

Allen, N.B., Badcock, P.B., 2006. Darwinian models of depression: a review of

evolutionary accounts of mood and mood disorders. Prog.

Neuropsychopharmacol. Biol. Psychiatry 30, 815–826.

American Psychiatric Association, 2000. Diagnostic and statistical manual of mental

disorders text revision (4th ed.). American Psychiatric Association, Washington,

DC.

Andréasson, A., Arborelius, L., Erlanson-Albertsson, C., Lekander, M., 2007. A

putative role for cytokines in the impaired appetite in depression. Brain

Behav. Immun. 21, 147–152.

Anyanwu, E., Campbell, A.W., Jones, J., Ehiri, J.E., Akpan, A.I., 2003. The neurological

significance of abnormal natural killer cell activity in chronic toxigenic mold

exposures. Sci. World J. 3, 1128–1137.

Ardayfio, P., Kim, K., 2006. Anxiogenic-like effect of chronic corticosterone in the

light-dark emergence task in mice. Behav. Neurosci. 120, 249–256.

Avery, D.H., Shah, S.H., Eder, D.N., Wildschiødtz, G., 1999. Nocturnal sweating andtemperature in depression. Acta Psychiatr. Scand. 100, 295–301.



12/14

Berk, L.S., Felten, D.L., Tan, S.A., Bittman, B.B., Westengard, J., 2001. Modulation of

neuroimmune parameters during the eustress of humorassociated mirthful

laughter. Alter. Ther. Health Med. 7 (62–72), 74–76.

Beseler, C.L., Stallones, L., 2008. A cohort study of pesticide poisoning and

dep

Documents

Depression as an Evolutionary Strategy for Defense Against Infection